使用Python计算与X公司的雇员任期

Question

我有一种情况，需要计算X年公司的雇员任期。数据以以下格式提供：

df =

EID      Event_Name                Event_Date
1        Hired                     03/09/1990  00:00:00
1        Terminated                14/10/2005  00:00:00
1        Rehired                   02/11/2015  00:00:00
2        Hired                     03/10/1990  00:00:00
2        Terminated                15/10/2005  00:00:00

权数计算很容易理解：

首先计算雇用和终止之间的差异秒计算今天的日期和重新雇用之间的差额（如果重新雇用则适用其他明智的忽略方法）

示例：对于EID = 1：任期为：

  (LAST JOB WORKED WITH COMPANY X)         14/10/2005  00:00:00 - 03/09/1990  00:00:00 = 5520 days
  (REHIRED WITH COMP X AND STILL WORKING)  25/04/2020  00:00:00 - 02/11/2015  00:00:00 = 1636 days

总天数=（5520 + 1636）/ 365 = 19.6年同样，对于其他EID也是如此

输出应如下所示：

 EID      Tenure(Years)               
 1        19.6
 2        15.04

Answer 1

如果Terminated列中每组从不连续的Event_Name值，则解决方案有效：

#converting to datetimes
df['Event_Date'] = pd.to_datetime(df['Event_Date'], dayfirst=True)

#today datetime
now = pd.Timestamp.now().floor('d')
#shifted rows per groups for terminated Event_Date 
df['new'] = df.groupby('EID')['Event_Date'].shift(-1).fillna(now)
#removed rows with Terminated
df = df[df['Event_Name'].ne('Terminated')].copy() 
#difference
df['Tenure(Years)'] = df['new'].sub(df['Event_Date'])
#aggregate sum
df = df.groupby('EID')['Tenure(Years)'].sum().dt.days.div(365).reset_index()
print (df)
   EID  Tenure(Years)
0    1      19.605479
1    2      15.043836